Derek Chun

Research project title: 
Application of energy based control algorithms on soft robotic systems.
Principal goal for project: 
The use of an energy based control system could enable higher efficiency based on the systems emergent dynamics compared to programmed kinematics.
Research project: 

My PhD project is based on a paper by Ross et al. (2016), “Controlling and Simulating Soft Robotic Systems: Insights from a Thermodynamic Perspective.” The paper discussed the potential of using energy based control algorithms on complex soft robots and the impact that simulation could have on soft robots development.

My project aims to demonstrate the theories (Hamiltonian and Thermodynamics) outlined by the paper on soft robots. Firstly, the control of physical robots with an energy based approach and secondly, the demonstration of a concurrent development process with simulation iterations to optimise physical soft robots.

Bio-inspired evolutionary algorithms will be important to enable the soft system to maximise its efficiency and perform the required task. 

The key areas which I am interested in are the following:-

  1. Thermodynamics - multi-domain physics understanding
  2. Dynamic simulation in particular Hardware In Loop (HIL) simulations
  3. Finite element analysis (FEA) for large scale deformation and variable stiffness
  4. Material science 
  5. Evolutionary algorithms / reinforcement learning to derive behaviour


About me: 

Before the CDT, I worked for AgustaWestland, the Anglo-Italian helicopter company for 5 years. Experience was gained in a variety of departments from Aircraft Operations, Transmissions Testing, Customer Training and Advanced Engineering (AW159 simulator), Customer Support and Procurement. Extensive experience in Programme Management of AW101 platform from bidding, building, aircraft delivery and customer support. 

During my MEng in Automotive Engineering at Loughborough University, I spent the year in industry at Honda Racing F1 systems working group. I was introduced to dynamic simlation of the driveline and data support at the factory for every races. The highlight was 3rd place finished by Rubens at the 2008 British GP.